Is Davos More Than a Boondoggle?...Is Davos More Than a Boondoggle? Andreas Fuchs Andrew K. Rosey...

Is Davos More Than a Boondoggle?

Andreas Fuchs∗ Andrew K. Rose†

Sebastian S. Schmidt‡

February 18, 2020

Paper presented at the 13th Annual Conference on

The Political Economy of International Organization

Vancouver, Canada, February 20–22, 2020

Abstract

Each year since 1971, the World Economic Forum (WEF) Annual Meeting in Davos,

Switzerland, has attracted the leadership of global corporations, governmental and

non-governmental organizations, as well as other public figures. However, attending

the Davos summit is costly for companies with estimated costs of US$40,000 per

delegate. On the one hand, WEF attendance could generate a value added for

companies if it generates a business network and buys valuable political support.

On the other hand, it could be wasteful in the sense that it generates only private

benefits to the attendees themselves without measurable effects for their companies.

Our paper is the first to study whether companies draw economic benefits from

attending Davos. We introduce a novel database on WEF attendees over the 2009—

2018 period and match it with firm-level data on stock market performance and

corporate ratings. We then use fixed-effects estimations and—in a future version

of this paper—a synthetic control method to test whether companies present at

Davos perform better. Our donor pool of counterfactual companies consists of

MSCI-ACWI corporations from the same sector. Our preliminary findings do not

provide evidence that the most famous summit of global leaders creates direct value

to businesses.

JEL Codes: F53, G24, G32, G39, O19

Keywords: World Economic Forum, international organizations, business leaders, stock

markets, corporate ratings, summits

∗University of Goettingen & Kiel Institute for the World Economy†UC Berkeley & National University of Singapore‡University of Goettingen & Kiel Institute for the World Economy

Acknowledgements: We thank Giannis Hafizis, Franziska Hartung, Elena Quante, Laura

Mahoney, Samuel Siewers, and Keonhi Son for valuable research assistance.

2

Heading to Davos, Switzerland, to

meet with World and Business

Leaders and bring Good Policy

and additional Hundreds of

Billions of Dollars back to the

United States of America! We are

now NUMBER ONE in the

Universe, by FAR!!

US President Donald J. Trump

Twitter, January 21, 2020

1 Introduction

The World Economic Forum (WEF), a not-for-profit organization founded in 1971,

describes itself as “the International Organization for Public-Private Cooperation,” which

“engages the foremost political, business and other leaders of society to shape global,

regional and industry agendas.”1 Its flagship is its Annual Meeting in Davos-Klosters,

Switzerland, which regularly amasses about 3,000 participants in the Alps. It attracts

the leadership of global corporations, governmental and non-governmental organizations,

the media, as well as other public figures from arts, culture, and sports. Attending Davos

is very costly for companies. In 2014, for example, the estimated costs were US$40,000

per delegate (Armstrong and Kottasova, 2014). These costs may be considered relatively

small if WEF attendance generates a measurable value added for companies. For example,

attending Davos might help strengthen business networks which companies could benefit

from in the short and long term if they lead to business deals. The Annual Meeting

might also be used as valuable opportunity to garner political support from country

leaders and other government representatives, or to create goodwill among civil society

representatives. In the words of a correspondent at Davos, “the value of Davos to the folks

who pay these fees has very little to do with the ‘themes’ of the conference (‘sustainability,’

and so forth). [...] Davos is now primarily a huge, high-level business conference, in which

senior executives from the world’s largest companies take advantage of their physical

proximity to meet in person with partners and clients and would-be clients—meetings

that can end up being vastly more valuable than the price of admission” (Blodget, 2011).

Yet, there are reasons to doubt that attending the summit actually benefits the

companies in a measurable way. Instead, Davos attendance might be a rather wasteful

1According to the mission statement, “[t]he Forum strives in all its efforts to demonstrateentrepreneurship in the global public interest while upholding the highest standards of governance.”See https://www.weforum.org/about/world-economic-forum for the full mission statement (accessed21 September 2019).

1

https://www.weforum.org/about/world-economic-forum

activity from the companies’ perspective and akin to a “vacation.”2 Davos might generate

private benefits to the attendees themselves if they, for example, like mingling with

celebrities, enjoy skiing with like-minded people, or use it as a job-market opportunity to

promote their career prospects. Rather than generating profit for the companies, sending

employees to Davos could be a waste of resources and mostly a distraction.

In order to study whether it pays off for companies to attend Davos, we construct

a novel database on WEF attendees over the 2009–2018 period and match it with firm-

level data on stock market performance and credit ratings. If it pays off to attend Davos

and if the stock market is efficient, companies who send delegates to the WEF should

show a better stock market performance than those that are not represented. Likewise,

one would expect that companies that are represented at the WEF are considered as

being more creditworthy than their non-participating counterparts. We start by running

regressions with firm-, country-year- and sector-year-fixed effects to isolate the effect of

Davos attendance. This allows us to exploit variation in firm performance within firms

over time for years in which the company sends delegates to Davos and those where it

does not, controlled for country- and sector-specific characteristics over time. To address

remaining endogeneity concerns, we will also use a synthetic control method in a future

version of this paper to test whether companies present at Davos perform better than

when compared to a synthetic twin company. Our donor pool of counterfactual companies

consists of MSCI-ACWI corporations that are not part of the Davos meetings. Our

findings shed light on whether the most famous summit of global leaders creates value to

businesses or just wastes (tax) money.

This paper is the first to study whether companies draw measurable economic benefits

from attending Davos. As such, our paper is related to the literature on the economic

benefits of international organizations (e.g., Rose, 2004, 2005, Nitsch, 2007a, Buthe

and Milner, 2008, Egger and Larch, 2011, Dreher et al., 2015). Within this literature,

our paper is closely related to the scholarly work that uses stock market performance

and credit ratings as outcome variable (Dreher and Voigt, 2011, Moser and Rose, 2014,

Davies and Studnicka, 2018, Gehring and Lang, 2018). In contrast to most contributions

in this literature that have so far been heavily focused on cross-country panel data, our

paper studies the economic benefits of an international body at the firm level. We also

contribute to the burgeoning economic literature on corporate social responsibility (see,

for an overview Kitzmueller and Shimshack, 2012).

We proceed as follows. Section 2 provides an overview on the WEF Annual Meeting

2For example, Davos participant Henry Blodget describes the Davos summit as “[t]heSuper Bowl [o]f Schmoozing” (https://www.businessinsider.com/davos-parties-2015-1?r=DE&IR=T) and his colleague Taylor Nicole Rogers notes that “[t]he parties arejust as wild as you’d imagine” (https://www.businessinsider.de/international/what-davos-is-really-like-long-lines-high-security-badges-2020-1/?r=US&IR=T, bothaccessed 13 February 2020).

2

https://www.businessinsider.com/davos-parties-2015-1?r=DE&IR=T

https://www.businessinsider.com/davos-parties-2015-1?r=DE&IR=T

https://www.businessinsider.de/international/what-davos-is-really-like-long-lines-high-security-badges-2020-1/?r=US&IR=T

https://www.businessinsider.de/international/what-davos-is-really-like-long-lines-high-security-badges-2020-1/?r=US&IR=T

in Davos and introduces a database on Davos attendance. In Section 3, we present our

empirical strategy and data required for our econometric analyses. Section 4 discusses

our findings from fixed-effects regressions and – in a future version of this paper – the

synthetic control method. We conclude the paper in Section 5.

2 The Annual Meeting of the World Economic

Forum

2.1 Background

The history of the WEF dates back to 1971, when it was founded as European

Management Forum. Back then, its main purpose was to provide a forum to discuss

how European companies could clear the backlog with respect to their management

practices in comparison to the United States. It is has since been developed into a

forum to discuss global challenges. To provide examples, the 2019 Annual Meeting was

held under the theme “Globalization 4.0: Shaping a Global Architecture in the Age

of the Fourth Industrial Revolution,” while the 2020 edition was labeled “Stakeholders

for a Cohesive and Sustainable World.” The Forum describes itself as being rooted in

stakeholder theory. Corporations, and civil society actors more broadly, are invited to

fill the void left by governments and international governmental organizations in global

governance. As the WEF founder and Executive Chairman Klaus Schwab puts it, the

world lacks “authentic and effective global leadership” and, as a result, “the influence of

corporations on communities, on the lives of citizens, and on the environment has sharply

increased” (Schwab, 2008, p.108). By participating in the Davos Annual Meetings,

attendees supposedly contribute to a global public good. The WEF itself describes its

work as “efforts to demonstrate entrepreneurship in the global public interest.”

The lion’s share of the participants come from the private sector.3 Beyond the

contribution to a global public good, rational companies are likely to seek private benefits

from attending as well. Since a four-digit number of other business leaders are present,

the Davos summit offers possibilities to make business deals with companies from every

continent. We thus hypothesize that companies perform better if they are represented at

Davos, all else being equal.

Moreover, since the Davos summit attracts an impressive number of political leaders,

it provides a unique opportunity to lobby according to business interests. US President

Donald Trump’s visit at the 2020 WEF offers a case in point. Among others, he met

with 15 European corporate leaders. In the media, companies that were represented

at this meeting were highlighted in media reports. For example, the pharmaceutical

3See WEF website at https://www.weforum.org/agenda/2019/01/

everything-you-need-to-know-about-davos-2019/, accessed 19 September 2019.

3

https://www.weforum.org/agenda/2019/01/everything-you-need-to-know-about-davos-2019/

https://www.weforum.org/agenda/2019/01/everything-you-need-to-know-about-davos-2019/

news website FiercePharma highlighted those companies in the pharmaceutical sector

that got invited by the US President in his apparent attempt to “try and gin up more

U.S. investment” and titled that “Novartis, Bayer CEOs get time with Trump as he

meets with EU business leaders during Davos trip.”4 Previous research suggests that

meetings with foreign leaders can promote economic exchange. Estimating export flows

from France, Germany and the United States over the 1948–2003 period, Nitsch (2007b)

shows that one state visit is associated with an eight-to-ten percent increase in exports.

Summarizing the findings of 32 prior contributions in a meta analysis, Moons and van

Bergeijk (2017) speak of an overall significantly positive effect of economic diplomacy on

trade and investment. It thus appears plausible that get-togethers with political leaders

at Davos promote business opportunities in a similar manner. We thus hypothesize that

the positive effect of Davos attendance on company performance is more pronounced in

years in which more country leaders are present, all else being equal.

At the same time, there are reasons to doubt that attending the Davos summit actually

provides private benefits to the attending companies in a measurable way. Instead,

attending Davos might be a rather wasteful venture and more akin to sending their

delegates on “vacation.” This has to be considered against the background of the high

fees that companies have to pay. The WEF levies annual membership and partnership

fees between CHF60,000 and CHF600,000 per company depending on their respective

membership type. In addition, companies need to pay a fee for each participant, which

was US$40,000 in 2014 (Armstrong and Kottasova, 2014).5 Any private benefit thus

needs to be traded off against these substantial costs.

Davos might only generate private benefits to the attendees themselves if they, for

example, like mingling with celebrities, enjoy skiing with like-minded people, or use it

as a job-market opportunity to promote their career prospects. A better positioning of

its corporate leadership on the job market might or might not align with a company’s

interests.

The Annual Meeting—and the anti-globalization protests that it attracts—also comes

with huge costs for the Swiss taxpayers6. Moreover, it is said to negatively influence the

quality of live of inhabitants and tourists in the Davos area. Survey evidence in Erfurt

and Johnsen (2003) shows that interviewees develop a more unpleasant view of the tourist

destination Davos when they are exposed to the WEF.

4See https://www.fiercepharma.com/pharma/during-davos-trip-trump-takes-meeting-novartis-bayer-ceos-and-other-execs(accessed 13 February 2020).

5Participants from the public sector and civil society organization typically do not pay fees.Travel grants are available for some participants such as academics. See WEF website at https:

//www.weforum.org/agenda/2017/01/who-pays-for-davos/, accessed 19 September 2019.6According to the public Swiss Radio and Television, the Swiss government spent CHF45

million in public security expenses for the 2020 summit (https://www.srf.ch/news/schweiz/sicherheitskosten-am-wef-polizeidirektoren-rechtfertigen-politischen-preis, accessed on15 February 2020).

4

https://www.fiercepharma.com/pharma/during-davos-trip-trump-takes-meeting-novartis-bayer-ceos-and-other-execs

https://www.weforum.org/agenda/2017/01/who-pays-for-davos/

https://www.weforum.org/agenda/2017/01/who-pays-for-davos/

https://www.srf.ch/news/schweiz/sicherheitskosten-am-wef-polizeidirektoren-rechtfertigen-politischen-preis

https://www.srf.ch/news/schweiz/sicherheitskosten-am-wef-polizeidirektoren-rechtfertigen-politischen-preis

2.2 Measuring Davos Attendance

We assemble the list of official attendees of the annual summit in Davos from several

sources. For the years 2009 to 2012 and 2014 to 2016, we use official attendee lists that

are available online. We append our data for the remaining years with reports from the

business news organization Quartz for the years 2013, 2017, and 2018.7 The data includes

the attendee’s name, organization, current position in the sending organization, as well

as the organization’s country of registration.

The resulting data set required substantial cleaning. The organization name was not

consistent across participants and years. For example, companies merge and split and the

delegates report only a division or subsidiary of a larger parent company. We cleaned this

variable to be able to track an organization’s participation over time. This included the

re-coding of nearly 900 organizations to its respective parent company. A prominent

example in the data set is Google. We reallocated Google attendees to its current

parent organization, Alphabet. Similarly, government institutions were registered in an

inconsistent way. This is why we grouped all attendees from governmental institutions

to their respective national or sub-national government. Third, we excluded delegates

sent by the WEF itself (e.g., staff members) and delegates who lack an affiliation, such

as self-employed individuals. Our final database includes 11,007 individuals from 4,910

individual organizations. For example, prominent participants in the 2018 summit, the

last summit covered in our dataset, include US President Donald Trump and the CEOs

of IBM, Virginia Rometty, and IKEA, Jesper Brodin, among many others.

The size of the Davos summit shows an upward trend. The total number of yearly

attendees increased from 2,282 in 2009 to 2,957 in 2018. Overall, attendees from 155

countries are represented. As highlighted in Figure 1, most participants come from the

United States (7,261), followed by the United Kingdom (2,847), and WEF home country

Switzerland (2,073). Also, the number of participating organizations increased from 1,469

to 1,667. On average, 382 new organizations, in the sense that they had not attended

an earlier summit in the observed period, attended Davos each year. In 2018 alone, 376

organizations joined, including the retail giant Walmart and the IT company NEC. The

company that sent most delegates in the observed period was Thomson Reuters (50),

followed by Bloomberg (45), and PepsiCo (36). The national government that sent the

most delegates was the United States (167), followed by the United Arab Emirates (104),

and Switzerland (71). The Annual Meeting is still dominated by high-income countries.

While the number of attendees from low- and middle-income countries has increased from

454 in 2009 to 712 in 2018, the share of attendees from this country group is stagnating

(see Figure B3 in the Appendix).

Next, we coded whether the organization is is part of the MSCI-All-Country-World-

7We thank David Yanofsky for kindly sharing his data with us.

5

Figure 1 – Total Number of Delegates by Country, 2009–2018

1 to 55 to 1010 to 2020 to 5050 to 100100 to 250250 to 500500 to 1,0001,000 to 2,0002,000 or more0

Source: Own visualization.

Figure 2 – Number of Delegates by Country Sent from ACWI-listed Corporations, 2009–2018

1 to 1010 to 5050 to 100100 to 500500 to 1,0001,000 or more0


Index (MSCI-ACWI). This market cap-weighted stock market index covers 2,746 stocks

from companies throughout the world. The MSCI-ACWI extends the more well known

MSCI-World by publicly listed firms from 26 emerging economies (MSCI, 2019).8 By

MSCI’s own account, the MSCI-ACWI covers around 85 percent of the free float-adjusted

market capitalization in each of the 47 national markets (MSCI, 2019). As can be seen

from Figure 2, most corporate participants at Davos (from MSCI-ACWI-listed companies)

8The MSCI-ACWI includes the following countries and territories: Australia, Austria, Belgium,Brazil, Canada, Chile, China, Colombia, Czech Republic, Denmark, Egypt, Finland, France, Germany,Greece, Hong Kong (China), Hungary, India, Indonesia, Ireland, Israel, Italy, Japan, South Korea,Malaysia, Mexico, Netherlands, New Zealand, Norway, Pakistan, Peru, Philippines, Poland, Portugal,Qatar, Russia, Singapore, South Africa, Spain, Sweden, Switzerland, Taiwan (China), Thailand, theUnited Kingdom, the United States, the United Arab Emirates, and Turkey.

6

come from the United States (1,985), followed by the United Kingdom (578), and India

(475).

Finally, to get an impression of the sectoral composition of these corporate delegates,

we group the companies into ten broad sectors as defined by Thomson Reuters (Reuters,

2012). As displayed in Figure 3, the largest share of attendees (24.4%) comes from the

sector Financials, followed by Industrials (12.9%) and Technology (11.5%).

Figure 3 – Composition of Corporate Delegates by Sector over Time, 2009–2018

0%

25%

50%

75%

100%

2009 2011 2013 2015 2017

Com

pani

es

Sector

Basic Materials

Consumer Cyclicals

Consumer Non−Cyclicals

Energy

Financials

Healthcare

Industrials

Technology

Telecommunication Services

Utilities


3 Empirical Strategy

3.1 Dependent Variables

Our goal is to investigate whether sending a delegate to the Davos summit affects firm

performance. To do so, we investigate the effect of Davos attendance on stock market

performance and corporate ratings.

3.1.1 Stock Market Performance

We follow previous research and use the MSCI-ACWI to map global stock market

performance (e.g., Moser and Rose, 2014). We collected data via Datastream for over

7

99% of all the 2,746 listed constituents.9 In very few cases, the MSCI-ACWI lists more

than one stock from a particular company. In these cases, we choose the share with

regular voting rights.10 In total, the stocks of relevance are traded at 54 national stock

markets from 51 countries and territories. For simplicity, we take the country in which

the stock is traded as the company’s country of origin. For 177 firms, however, these

two characteristics differ. We use the country as listed in Datastream and only deviate

from this in some specific cases (e.g., countries defined as tax havens) where we use the

country of the operational headquarter instead. We identify the operational headquarter

by Internet research.

Our measure of stock market performance is the end-of-the-year excess return for each

of the 2,727 identified constituents.11 We download this variable in their respective local

currencies for the years 2000 to 2018. We then convert all values into US dollar using daily

closing spot exchange rates, which we also collect via Datastream, Thomson Reuters,

GTIS, and MSCI. To control for the overall market movement, we regress the firms’

end-of-the year returns on the end-of-the-year market return of the MSCI-ACWI and

the respective intercept in order to estimate the firm specific beta following MacKinlay

(1997). This way, we control for the co-movement of a firm’s return and the return of the

MSCI-ACWI as well as any possible deviation of the firm compared to the index (Moser

and Rose, 2014). We use the annual market return on the MSCI-ACWI from 2008 to

2018.

As a next step, we match the list of companies listed in the MSCI-ACWI with our

data set of Davos attendees and identify the 516 MSCI-ACWI companies that attended

the WEF Annual Meeting at least once in the observed period. This corresponds to 18.7%

of all companies listed in the MSCI-ACWI. These 516 companies account for 42.2% of

the index’s capitalization. This shows that many globally active corporations do not

attend the summit, which will allow us to build a meaningful counterfactual. Apple, for

example, does not attend any of the summits in our sample and accounts for more than

2% of the total capitalization of the MSCI-ACWI index.

3.1.2 Corporate Credit Ratings

We also study the effect of Davos attendance on the credit ratings of companies. By

doing so, we intend to learn whether Davos attendance affects a more long-term indicator

of firm performance. Specifically, we use the long-term foreign-currency issuer credit

ratings at the end of the respective year of the three largest credit rating agencies: Fitch,

Moody’s, and Standard & Poor’s. We retrieve the data via Datastream and convert the

9The list of constituents was accessed at the MSCI website on 23 April 2019 (https://www.msci.com/constituents).

10Examples in our dataset are: “Alphabet A” and “Alphabet C,” or “Volkswagen Stamm” and“Volkswagen Vorzug,” where we exclude “Alphabet C” and “Volkswagen Vorzug.”

11We define the end of the year as the last trading day of the particular stock market in a given year.

8

https://www.msci.com/constituents

https://www.msci.com/constituents

alphanumeric rating levels into a uniform numerical scale, where 1 is the lowest rating

category (rating “C” or lower) and 21 is the possible best possible outcome (“AAA” for

Fitch and Standard & Poor’s and “Aaa” for Moody’s). This approach is standard in the

literature on credit ratings (Fuchs and Gehring, 2017). If the credit rating differs across

agencies, we take the average value among those agencies that issues a rating for the

respective company.

The resulting estimation sample contains an unbalanced panel of credit ratings for

1,543 firms, of which 399 companies have at least once been represented at the WEF.

3.2 Fixed-effects Regressions

To estimate the effect of attending Davos on firm performance, we estimate the following

equation with least squares:

Rijst = βDavosijst + φijs + σst + ξjt + εijst,

where Rijst is the end-of-the-year excess return or corporate rating in year t of company

i, which is located in country j and belongs to sector s; Davosijst is a binary variable that

takes a value of one if company i has at least one participant at the Davos summit in year

t; φijs are firm-fixed effects; and σst and ξjt denote sector-year- and country-year-fixed

effects, respectively. The inclusion of firm-fixed effects allows us to control for all time-

invariant firm-specific characteristics such as the sector, location, and corporate history.

By adding sector-year-fixed effects, we control for time-varying factors that affect an

entire sector such as technological progress.12 The inclusion of country-year-fixed effects

enables us to further control for time-specific factors that affect an entire country such as

a change of government policies or an economic downturn. Standard errors are clustered

at the firm level.

It is not clear whether WEF membership or actual attendance at the Davos summit

is the more appropriate representation of our variable of interest. However, the WEF

does not provide a comprehensive list of all members.13 Since every for-profit company

that attends the Davos summit must be a WEF member, we believe that attendance

is also a good proxy for membership. Moreover, there are reasons to expect that the

number of attendees matters in addition to participation. Companies pay for premium

membership to be able to send additional delegates, supposedly to have more workforce

to make business deals and lobby politicians. We repeat the regression from above but

replace the attendance dummy with the actual number of attendees sent by a company

12We group the companies in ten sectors: Basic Materials, Consumer Cyclicals, Consumer Non-Cyclicals, Energy, Financials, Healthcare, Industrials, Technology, Telecommunication Services, andUtilities.

13More than 1,000 companies are WEF members. See https://www.weforum.org/about/

our-members-and-partners.

9

https://www.weforum.org/about/our-members-and-partners

https://www.weforum.org/about/our-members-and-partners

in a given year.

Table 1 provides descriptive statistics on all variables employed in our paper. We

show separate statistics for our estimation sample for the return (panel A) and the rating

(panel B) regressions. Focusing our interpretation on the larger sample presented in

panel A, we observe that 10% of all company-years participated at the Davos summit.

The maximum number of attendees sent to Davos was 15. Each year 1.1% of all companies

are “newcomers” at Davos in the sense that they had not been represented at a previous

summit in our sample.

Table 1 – Summary Statistics

count mean sd min max

Panel A: Returns

Returns 25264 0.000 0.474 -1.314 13.794Company attendance 25264 0.100 0.301 0.000 1.000No. of delegates 25264 0.236 0.908 0.000 15.000Attendance size 2539 2.345 1.807 1.000 15.000Lagged Outlook 12769 1.929 0.473 1.000 3.000Lagged Rating 10791 13.646 2.676 1.000 21.000Lagged Return 22538 0.001 0.494 -1.314 13.794G20 Gov’t officials 25264 27.002 5.327 19.000 37.000G20 Leaders 25264 5.814 2.016 3.000 9.000G20 Leaders/GDP 25264 0.253 0.151 0.061 0.587MSCI World 25264 0.621 0.485 0.000 1.000Newcomer 22966 0.011 0.104 0.000 1.000

Panel B: Ratings

Credit Rating 11535 13.609 2.650 1.000 21.000Company attendance 11535 0.156 0.363 0.000 1.000No. of delegates 11535 0.382 1.152 0.000 15.000Attendance size 1805 2.443 1.858 1.000 15.000Lagged Outlook 10892 1.915 0.481 1.000 3.000Lagged Rating 11051 13.648 2.699 1.000 21.000Lagged Return 10071 -0.037 0.380 -1.262 5.454G20 Gov’t officials 11535 27.232 5.351 19.000 37.000G20 Leaders 11535 5.836 2.045 3.000 9.000G20 Leaders/GDP 11535 0.255 0.155 0.061 0.587MSCI World 11535 0.773 0.419 0.000 1.000Newcomer 10576 0.017 0.128 0.000 1.000

Note: Descriptive statistics based on sample used in Table 2, column 1.

3.3 Synthetic Control Method

In a later version of this paper, we will apply a synthetic control method, which is a

popular method to analyze causal dependencies in observational data. Previous research

shows that this method yields a consistent estimator for non-randomized treatment effects

and this even holds under multiple treatments and points in time of treatment (Abadie

and Gardeazabal, 2003, Abadie et al., 2010, Acemoglu et al., 2016, Xu, 2017).

10

In the synthetic control method, the treatment effect is the difference between the

treated observation and its synthetic control after the point of treatment (Acemoglu

et al., 2016). For each observation in the treatment group we will create a counterfactual

by drawing a weighted average from all observations in the control group so that the

treatment and control are as similar as possible to results prior to the treatment. Our

donor pool of counterfactual companies will consist of MSCI-ACWI corporations from

the same sector that do not participate at the annual meeting. As a pre-treatment period,

we will use the ten years prior to the first time a company sends a delegate to the Davos

summit to avoid unobserved pre-treatment effects. We will then test the validity of our

estimated treatment effect in a placebo test, where we randomly assign treatment to a

corporation from the donor pool. To avoid a cumbersome repetition of the methodology

for each treated corporation, we will apply the generalized synthetic control method

proposed by (Xu, 2017).

4 Results

Table 2 shows our baseline results. Column 1 includes firm- and year-fixed effects only;

column 2 replaces year-fixed effects by the stricter country-year-fixed effects, column

3 replaces year-fixed effects by the stricter sector-year-fixed effects, and column 4 shows

results with both country-year-fixed effects and sector-year-fixed effects. Panel A presents

the results for excess returns. Rather than showing evidence for beneficial effects on

firm performance, the coefficient on company attendance at Davos is even negative

in all four columns. However, the coefficients do not reach statistical significance at

conventional levels. This is evidence against the hypothesis that companies benefit from

Davos attendance in terms of stock market performance.

Being a Davos participant could be perceived as a stronger signal in emerging markets

compared to companies in advanced economies. To test this, the regression in column

5 restricts the sample to companies based in one of the 26 emerging economies. The

sample is significantly smaller and covers 1,050 firms only. However, this company

group is particularly dynamic and we should observe more “newcomers” at the WEF

here. However, in contrast to our expectations, the coefficient on our variable of interest

even becomes larger in absolute size and becomes statistically significant at the ten-

percent level. We conclude that firms participating in Davos do not even show a better

performance for this most-likely group of countries.

In panel B, we replicate our analysis with credit ratings as our dependent variable. The

coefficients on Davos attendance are economically and statistically insignificant. Together

with our previous finding for returns, we tentatively conclude that it does not pay off for

companies to attend Davos.

So far, each company entered our regression with the same weight. Next, we test

11

Table 2 – Baseline Regression Results, OLS, 2009–2018

(1) (2) (3) (4) (5)EmergingMarkets

Panel A: Returns

Company attendance -0.025 -0.025 -0.019 -0.018 -0.055*(0.018) (0.017) (0.017) (0.016) (0.031)

Observations 25264 25264 25264 25264 9570R2 0.026 0.153 0.054 0.178 0.102Firms 2726 2726 2726 2726 1050

Panel B: Ratings

Company attendance 0.015 0.018 0.017 0.021 -0.106(0.060) (0.060) (0.058) (0.058) (0.102)


Country-Year FE YES YESSector-Year FE YES YES

Notes: The dependent variable in panel A is annual excess returns of shares included in the MSCI-ACWI which are calculated as the relative change of the end-of-the-year closing rate measured inUS $. The dependent variable in Panel B is long-term issuer ratings. All specifications include aconstant, firm and year fixed effects. Clustered standard errors in parentheses. *** p<0.01, **p<0.05, * p<0.1.

whether our results hold, when we run weighted regressions. We weight companies

by their respective market value from the previous year, normalized to one, over the

cumulative market value of the MSCI-ACWI also from the previous year. As can be seen

in Table 3, we again observe no significant coefficient in any of the five regressions in both

panels. This increases our confidence that there is indeed an average null effect of Davos

attendance on firm performance.

In Table 4, we replicate our previous regressions but replace the simple Davos dummy

with the number of company attendees at Davos in a given year. It might be that firm

performance is only affected by a larger number of delegates. By replacing our binary

variable with a count variable, we can account for this. However, as our regression results

show, we again do not find evidence of a positive Davos effect on firm performance in any

of the five specifications and in both panels.

Next, we control for a special effect of the first year of being at Davos. It might

be that new participants attract particular attention, which improves firm performance

in the first year. The regression specifications in Table 5 include an additional binary

variable that takes a value of one if a company is a first-time participant in a given year.

As the results show, there is also no Davos effect neither on excess returns nor credit

12

Table 3 – Weighted Regression Results, OLS, 2009–2018


Panel A: Returns

Company attendance -0.025 -0.025 -0.019 -0.018 -0.055*(0.019) (0.018) (0.018) (0.017) (0.033)


Panel B: Ratings

Company attendance 0.015 0.018 0.017 0.021 -0.106(0.064) (0.064) (0.062) (0.062) (0.110)



Notes: The dependent variable in panel A is annual excess returns of shares included in the MSCI-ACWI which are calculated as the relative change of the end-of-the-year closing rate measured inUS $. The dependent variable in Panel B is long-term issuer ratings. All specifications include aconstant, firm and year fixed effects. Weights are lagged end-of-year market values of all MSCI-ACWI constituents normalised to one. Clustered standard errors in parentheses. *** p<0.01, **p<0.05, * p<0.1.

ratings for first-time participants.

Our non-finding could be driven by a lack of control for selection into the group of

WEF companies. While we cannot directly for selection, it is unlikely that our null

finding is driven by selection bias, as more rather than less performing firms should be

more likely to become WEF members. What is more, we run a robustness test that is

not subject to a selection effect. Specifically, we only analyze the potential effect of the

number of delegates a company has at Davos on firm performance within the group of

participants. If Davos attendance improves firm performance, we would also expect that

stronger participation translates into better performance and this should be visible in

this reduced sample that cannot be subject to the aforementioned sample selection bias.

As presented in detail in Table A4 in the Appendix, there is also no evidence that Davos

attendance pays off for companies within the group of WEF companies.

Finally, it could be that the effect is only visible in those years in which a large number

of high-ranked politicians attend the Annual Meeting. This is why we also identify how

many important government officials from the G20 member states attend the summit in

a given year. As important government officials, we code the head of state, the head of

13

Table 4 – Delegate Count Regression Results, OLS, 2009–2018


Panel A: Returns

No. of delegates -0.002 -0.004 -0.001 -0.002 -0.022(0.011) (0.010) (0.010) (0.010) (0.018)


Panel B: Ratings

No. of delegates 0.004 0.002 0.006 0.006 -0.050(0.029) (0.028) (0.029) (0.028) (0.057)




government, the minister of economic affairs, the minister of finance, and the minister of

foreign affairs. The number of high-level politicians shows considerable variation over time

and ranges between 20 (2011) and 38 (2016) (see Figures B1 and B2 in the Appendix). To

test this, we interact our Davos dummy with the number of important G20 government

officials present at Davos. The upper panel of Figure B4 shows the marginal effects

of Davos attendance for all values of G20 government official in our sample. Even in

important years there does not appear to be a statistically significant effect on firm

performance. In the lower panel of Figure B4, we replicate the results when we focus on

the G20 heads of state or head of government being represented rather than the total

number of all important G20 government members. Again, Davos attendance does not

positively affect firm performance. For very low values of G20 heads we even observe

significant negative effects. Summing up, we could not detect a measurable robust effect

of Davos attendance on firm performance.

14

Table 5 – Regression Results with Newcomers, OLS, 2009–2018


Panel A: Returns

Company attendance -0.020 -0.016 -0.017 -0.012 -0.029(0.014) (0.013) (0.014) (0.013) (0.031)

Newcomer 0.012 0.016 0.015 0.017 -0.006(0.024) (0.023) (0.023) (0.022) (0.052)


Panel B: Ratings

Company attendance -0.013 -0.009 -0.016 -0.009 -0.113(0.067) (0.067) (0.065) (0.065) (0.112)

Newcomer 0.076 0.038 0.105 0.066 0.096(0.068) (0.067) (0.068) (0.066) (0.099)




5 Concluding Remarks

Corporate leaders have high opportunity costs of attending a summit. We investigate

the corporate benefits from attending the Annual Summit of the World Economic

Forum in Davos, Switzerland, which attracts hundreds of business leaders. To study

empirically whether WEF participation pays off, we introduced a novel database on

WEF attendees over the 2009 to 2018 period and matched it with firm-level data on stock

market performance as well as credit ratings. We then used fixed-effects estimations to

test whether the companies present at Davos perform better and whether the number

of attendees matters once a company participates. Our preliminary findings do not

provide evidence that the most famous summit of global leaders creates a direct value to

businesses.

It is unlikely that this null finding is driven by reverse causality, as more rather than

less performing firms should be more likely to become WEF members (if at all). Reversing

15

the causal link and estimating a regression specification that explains entry into the WEF,

it also does not appear that our results are subject to reverse-causality bias (see Figure 5

for details). In a future version of this paper, we will also use a synthetic control method

to address remaining endogeneity concerns. Specifically, we will test whether companies

present at Davos perform better than when compared to a synthetic twin company.

Does this mean that Davos is just a boondoggle? Our paper investigates whether

Davos attendance is beneficial to the companies themselves. Taking the perspective of

profit-maximizing companies, our tentative answer is yes. However, we also acknowledge

that our study is only a first step towards a better understanding of the effects of one of

the world’s most prestigious summits. Future research should study the effects of Davos

attendance from the perspective of delegates and global citizens. First, it is of interest

whether employees benefit from Davos attendance, e.g., in terms of career prospects,

salary, location in business networks, and job satisfaction. Second, scholars could

investigate whether the agenda of the Davos summit shapes public discourses.14 Although

our study reveals null effects from a firm perspective, companies might contribute to a

global public good that significantly shapes global agendas. For example, survey data

may enable researchers to study whether media coverage around the Davos summit shapes

public opinion in countries with strong representation. Such an analysis would help

evaluate whether the Davos summit indeed contributes to a global public good.

14Based on interviews with Davos attendees, Giesler and Veresiu (2014) study the effects of the WEF’sproblem-solving initiatives on consumers. They conclude that the WEF is not “improving the state ofthe world” but rather reinforcing the global problems by shifting the burden on responsible consumers.

16

References

Abadie, A., Diamond, A. and Hainmueller, J. (2010), ‘Synthetic control methods

for comparative case studies: Estimating the effect of California’s tobacco control

program’, Journal of the American Statistical Association 105(490), 493–505.

Abadie, A. and Gardeazabal, J. (2003), ‘The economic costs of conflict: A case study of

the Basque country’, American Economic Review 93(1), 113–132.

Acemoglu, D., Johnson, S., Kermani, A., Kwak, J. and Mitton, T. (2016), ‘The value of

connections in turbulent times: Evidence from the United States’, Journal of Financial

Economics 121(2), 368–391.

Armstrong, P. and Kottasova, I. (2014), ‘Davos: All you need to know to be a delegate’.

CNN.

URL: http://edition.cnn.com/2014/01/15/business/davos-wef-delegate-

guide/index.html

Blodget, H. (2011), ‘The truth about Davos: Here’s why people happily pay $71,000 to

come–and why they’ll keep paying more every year’. Business Insider.

URL: https://www.businessinsider.com/costs-of-davos-2011-1?IR=T

Buthe, T. and Milner, H. V. (2008), ‘The politics of foreign direct investment

into developing countries: Increasing FDI through international trade agreements?’,

American Journal of Political Science 52(4), 741–762.

Davies, R. B. and Studnicka, Z. (2018), ‘The heterogeneous impact of Brexit: Early

indications from the FTSE’, European Economic Review 110, 1–17.

Dreher, A., Mikosch, H. and Voigt, S. (2015), ‘Membership has its privileges–the effect of

membership in international organizations on FDI’, World Development 66, 346–358.

Dreher, A. and Voigt, S. (2011), ‘Does membership in international organizations

increase governments’ credibility? testing the effects of delegating powers’, Journal

of Comparative Economics 39(3), 326–348.

Egger, P. and Larch, M. (2011), ‘An assessment of the Europe agreements’ effects on

bilateral trade, GDP, and welfare’, European Economic Review 55(2), 263–279.

Erfurt, R. A. and Johnsen, J. (2003), ‘Influence of an event on a destination’s

image—the case of the Annual Meeting of the World Economic Forum (WEF) in

Davos/Switzerland’, Tourism Review 58(4), 21–27.

Fuchs, A. and Gehring, K. (2017), ‘The home bias in sovereign ratings’, Journal of the

European Economic Association 15(6), 1386–1423.

Gehring, K. and Lang, V. (2018), ‘Stigma or cushion? IMF programs and sovereign

creditworthiness’. CESifo Working Paper No. 7339.

Giesler, M. and Veresiu, E. (2014), ‘Creating the responsible consumer: Moralistic

governance regimes and consumer subjectivity’, Journal of Consumer Research

17

41(3), 840–857.

URL: https://doi.org/10.1086/677842

Kitzmueller, M. and Shimshack, J. (2012), ‘Economic perspectives on corporate social

responsibility’, Journal of Economic Literature 50(1), 51–84.

MacKinlay, A. C. (1997), ‘Event studies in economics and finance’, Journal of economic

literature 35(1), 13–39.

Moons, S. J. and van Bergeijk, P. A. (2017), ‘Does economic diplomacy work? a meta-

analysis of its impact on trade and investment’, World Economy 40(2), 336–368.

Moser, C. and Rose, A. K. (2014), ‘Who benefits from regional trade agreements? the

view from the stock market’, European Economic Review 68, 31–47.

MSCI (2019), ‘MSCI ACWI Index Brochure’. .

Nitsch, V. (2007a), ‘Does the G7/G8 promote trade?’, Economics Letters 94(1), 136–140.

Nitsch, V. (2007b), ‘State visits and international trade’, World Economy 30(12), 1797–

1816.

Reuters, T. (2012), ‘Thomson Reuters Business Classification Methodology’.

Rose, A. (2005), ‘Which international institutions promote international trade?’, Review

of International Economics 13(4), 682–698.

Rose, A. K. (2004), ‘Do we really know that the WTO increases trade?’, American

Economic Review 94(1), 98–114.

Schwab, K. (2008), ‘Global corporate citizenship: Working with governments and civil

society’, Foreign Affairs 87(1), 107–118.

Xu, Y. (2017), ‘Generalized synthetic control method: Causal inference with interactive

fixed effects models’, Political Analysis 25(1), 57–76.

18

https://www.msci.com/documents/1296102/15035999/USLetter-MIS-ACWI-Apr2019-cbr-en.pdf/9de006ef-9cf0-8bd7-62ec-e67430e9155f?t=1559105406131

Appendix

A Additional tables

Table A1 – Conversion of credit ratings into numerical values (21-point scale)

Fitch Moody’s S&P Numerical scale

AAA Aaa AAA 21AA+ Aa1 AA+ 20AA Aa2 AA 19AA- Aa3 AA- 18A+ A1 A+ 17A A2 A 16A- A3 A- 15

BBB+ Baa1 BBB+ 14BBB Baa2 BBB 13BBB- Baa3 BBB- 12BB+ Ba1 BB+ 11BB Ba2 BB 10BB- Ba3 BB- 9B+ B1 B+ 8B B2 B 7B- B3 B- 6

CCC+ Caa1 CCC+ 5CCC Caa2 CCC 4CCC- Caa3 CCC- 3CC Ca CC 2C C C 1

DDD SD 1DD 1D D 1

RD 1

Source: Own adaption, based on Fuchs and Gehring (2017).

19

Table A2 – Size Effect of Attendees Groups, OLS, 2009–2018


Panel A: Returns

Attendance size 0.005 0.005 -0.003 -0.000 0.006(0.016) (0.013) (0.015) (0.014) (0.030)


Panel B: Ratings

Attendance size 0.029 0.014 0.048 0.039 -0.072(0.040) (0.040) (0.041) (0.041) (0.097)




Table A3 – Explaining Newcomers

(1) (2) (3) (4)

Lagged Return -0.044 -0.098 -0.050 -0.083(0.044) (0.069) (0.048) (0.075)

Lagged Rating 0.025 0.043 0.047 0.032(0.031) (0.061) (0.032) (0.055)

Lagged Outlook 0.064* 0.107** 0.048 0.110*(0.034) (0.050) (0.041) (0.060)

Observations 635 635 635 635R2 0.387 0.593 0.485 0.701Firms 151 151 151 151


Notes: The dependent variable is whether a firm is a newcomer. All specifications include aconstant, firm and year fixed effects. Clustered standard errors in parentheses. *** p<0.01, **p<0.05, * p<0.1.

20

Table A4 – Sources of WEF Annual Meeting Lists

Year Description Source

2009 World Economic Forum Annual Meeting List of Participants as of 30April 2013

Online


Online


Online


Online

2013 World Economic Forum Annual Meeting List of Participants Quartz.com2014 World Economic Forum Annual Meeting List of Participants as of 14

January 2014Online

2015 World Economic Forum Annual Meeting List of Participants as of 20January 2015

Online


Online


David Yanofsky


David Yanofsky

21

B Additional figures

Figure B1 – Total Number of G20 Leaders at the WEF Annual Meeting, 2009–2018

4 to 55 to 1010 to 1515 to 2020 to 2525 to 3030 to 34Not G20


Figure B2 – Total Number of G20 Leader from 2009 to 2018

0

10

20

30

2010 2013 2016Year

Num

ber

of G

20 L

eade

rs


22

Figure B3 – Share of Delegates Sent from Low- and Middle-income Countries, 2009–2018

0%

25%

50%

75%

100%

2010 2013 2016Year

Sha

re o

f del

egat

es

High Income

Low & Middle Income


Figure B4 – Marginal effects of regressions with interaction term, OLS, 2009–2018

-.04

-.02

0.0

2Ef

fect

s on

Lin

ear P

redi

ctio

n

19 21 23 25 27 29 31 33 35 37G20 Gov’t officials

Average Marginal Effects of davos_year with 90% CIs

-.2-.1

5-.1

-.05

0.0

5Ef

fect

s on

Lin

ear P

redi

ctio

n

19 21 23 25 27 29 31 33 35 37G20 Gov’t officials


-.1-.0

50

.05

Effe

cts

on L

inea

r Pre

dict

ion

3 4 5 6 7 8 9G20 Leaders


-.2-.1

0.1

.2Ef

fect

s on

Lin

ear P

redi

ctio

n

3 4 5 6 7 8 9G20 Leaders


Notes: The figure displays marginal effects together with a 90% confidence interval. Thedependent variable in the left column is annual excess returns of shares included in the MSCI-ACWI which are calculated as the relative change of the end-of-the-year closing rate measured inUS $. The dependent variable in the right column is long-term issuer ratings. All specificationsinclude country-year- and sector-year-fixed effects.

23

Is Davos More Than a Boondoggle?...Is Davos More Than a Boondoggle? Andreas Fuchs Andrew K. Rosey...

Documents

Transcript of Is Davos More Than a Boondoggle?...Is Davos More Than a Boondoggle? Andreas Fuchs Andrew K. Rosey...